2.0 Analysis 2.1 Time of Flooding of the Shelter Deck There was no water on the shelter deck when the master of the MARWOOD visited it at about 2345, at which time the tide was still falling. Further, at about 0200, the junior deck-hand observed water being discharged from the overboard discharge for the shelter deck bilge pump. This would suggest that water had found its way onto the shelter deck sometime between 2345 and 0200, before the JO MARC sailed. Furthermore, being new to the vessel, the junior deck-hand did not realize the significance of his observation and, consequently, did not bring it to the attention of anyone else. 2.2 Action by the Owner of the MARWOOD At 0215, the deck-hands of both vessels, at the behest of the owner of the MARWOOD, checked the mooring lines to ensure that none were caught under the wharf fixtures. The owner of the MARWOOD, who was aware that the shelter deck occasionally flooded while at sea, was to sail on the JO MARC. However, the shelter deck was not inspected before the owner of the MARWOOD boarded the JO MARC, even though the practice on board was to leave the gate valve for the conveyor drain line open at all times. 2.2.1 Action by the Chief Engineer After docking, the chief engineer filled the starboard fresh-water tank, creating a five- to seven-degree starboard list, and then proceeded ashore at 2030. A similar list was observed by the owner at 2230 and by the master around midnight. Further, after ensuring that the mooring lines were not entangled with the wharf structure, both masters sailed aboard the JO MARC around 0215. The owner was not overly concerned about the list, which would indicate that it had not increased markedly. The fact that the filling valves for the starboard forward ballast/fish tank (13) and the port middle fish hold (16) were found open, and that the water level in the former was at a higher level than in the latter, would indicate that these valves had been opened to counter the effect of the starboard list and that the valves had been opened by the chief engineer following his return to the vessel. In addition, and based on the status of the electrical switches/systems, the position of the valves, and the operational mode in which the equipment was found upon salvage, it would appear that the chief engineer carried out a number of activities upon his return to the vessel, which included the following: he changed to work clothes; he shut the small bilge pump and started the large brine circulation/bilge pump to pump out water from the shelter deck; he closed the gate valve for the fish conveyor drain well on the shelter deck located in the galley dry store, a known source of water ingress to the shelter deck in the past. 2.3 Alcohol Consumption and Performance Although the chief engineer had consumed a quantity of alcohol which, according to blood alcohol measurements, would have put him under the influence, he nonetheless took a series of purposeful actions, based on his assessment of the situation, which was most likely influenced by his experience of past flooding. It is not known if the amount of alcohol consumed could have impaired his operational ability. The chief engineer's efforts to prevent the vessel from sinking were overcome by the rapidity of the developing situation. 2.4 Chief Engineer Trapped Below Deck When the chief engineer was in the dry store to shut the gate valve to the conveyor well, the vessel's list was continually increasing. A large starboard list would have shifted the consumables in the refrigerator. As the refrigerator door was held closed by a magnetic catch only, the weight of the consumables in and against the door would have permitted it to swing open. When the refrigerator door was opened, its handle would have become interlocked with the dry store door handle, effectively trapping the chief engineer inside the dry store. This is consistent with the circumstances in which the chief engineer's body was found. 2.5 Sequence of Events After being replenished with fresh water, the vessel had developed a list of five to seven degrees. The list increased the amount of overhang of the spare trawl doors on the starboard after main deck. The hazard associated with leaving a vessel unattended with such a list was not fully appreciated by the crew. Between the time when the chief engineer and the master proceeded ashore and the time when the master visited the shelter deck (when no water was noted), the tide was falling. By the time the crew returned to the vessel shortly before 0200, some water had to have found its way onto the shelter deck for the pump to be discharging full bore. At the time the vessel was left unattended, the tide was falling; having reached low water at about 0037, it was rising when the JO MARC sailed from port. With the mooring lines unattended and the vessel stemming the flood, the bow would be able to swing away from the wharf and the stern into the wharf. The actions of the chief engineer following his arrival on board at about 0245 would suggest that, by this time, the vessel's list had increased. This probably meant that the overhanging trawl doors, by then, were inextricably caught under the wale of the dock apron. The actions taken by the chief engineer were consistent with those required to rectify the list and address possible flooding from a known previous source of ingress. In any event, the situation had progressed to such an extent that it could not have been remedied by the chief engineer's action(s) alone. With the vessel restrained, the rising tide would have caused the vessel's list to starboard to increase. The open gate valve for the fish conveyor drain well and the non-watertight fish hold hatch covers would have allowed the water to accumulate rapidly onto the starboard side of the shelter deck. The cumulative effect of this additional weight further outboard would have exacerbated the starboard list. By the time the gate valve to the conveyor drain line was shut, the situation had deteriorated substantially. With the deck edge submerged and the watertight doors leading to the accommodation/engine-room being left open/not securely shut, rapid downflooding into those compartments ensued, further increasing the list. The downflooding continued until the vessel lost reserve buoyancy and sank. 2.6 Fish Conveyor Overboard Drain Although the overboard drain for the fish conveyor well was positioned below the shelter deck and was submerged even in the lightly loaded condition, the CCG SSB permitted the fitting of a gate valve and did not require the installation of a non-return valve. The hazard associated with such an arrangement is reflected in the MARWOOD experiencing flooding of the shelter deck. Further, during fish loading/processing, the overboard drain valve had to be kept open to allow water to be drained into the sea. This resulted in the practice, albeit unsafe, aboard the MARWOOD of keeping the valve open at all times. Although the MARWOOD was alongside the wharf, the valve was left open and contributed to the sinking of the vessel. The fitting of a non-return valve would have prevented ingress of water onto the shelter deck through this source. Furthermore, there is no regulatory requirement for fish holds converted for the carriage of fish in liquid to be treated as tanks. Hence, there is no requirement for increased bulkhead scantlings, for watertight tank lids instead of hatch covers, or for the holds to have vent pipes installed to relieve internal pressure or vacuum. 2.7 Hatch Covers During the conversion to a stern trawler, the original fish hold watertight hatch covers were replaced by light aluminium covers of a non-watertight design. While this is permitted under the LFVIR, it exacerbated the deteriorating stability condition by allowing the water in the fish holds to flow onto the shelter deck due to the large list. 2.8 Quality of Steamship Inspections As the MARWOOD had undergone modifications in 1987 which adversely affected her stability, the approved Stability Booklet was no longer valid. The vessel, therefore, had been operating without an approved Stability Booklet for over six years. The modifications, which included installation of the gantry and a large net drum, were obvious during inspections. Further, the presence of eight long tons of permanent ballast positioned in the shaft tunnel and the engine-room bilge in way of the gearbox should have been evident during quadrennial inspection. Although six annual inspections and one quadrennial inspection were carried out following the 1987 modifications, the CCG SSB reportedly only became aware of these modifications following this occurrence. As the main objective of steamship inspections is to ensure that vessels proceeding to sea are in a seaworthy condition, good inspection practice dictates that all aspects of seaworthiness should be considered when visiting a vessel for inspection. It is the responsibility of the owner or master of the vessel to notify the CCG SSB of any alteration which may affect compliance with the regulations. However, it is also incumbent upon the steamship inspector to be vigilant and to inquire of the master or owner if any alteration has been carried out between inspections. 3.0 Conclusions 3.1 Findings The vessel was left unattended at the wharf with a starboard list at a tidal port, and the associated hazard was not fully appreciated by the crew. The spare trawl doors which were overhanging the starboard ship side became caught under the wale of the wharf on a rising tide. Water ingress to the shelter deck was through the open gate valve for the conveyor well. The authorities permitted the discharge from the conveyor well to be below the deepest load waterline and the fitting of a gate valve. The vessel's shelter deck was known to flood through the overboard discharge line to the conveyor well; despite this, the practice of keeping the gate valve on the line open continued. With the vessel's starboard side caught under the wale of the wharf on the rising tide, the rate of ingress through the overboard discharge pipe to the conveyor well increased, thereby increasing the starboard list. With the vessel in a listed condition, the non-watertight hatch covers allowed water to spill onto the starboard shelter deck, further exacerbating the situation and frustrating the attempts of the chief engineer to remedy the situation. The chief engineer's action to remedy the situation was consistent with the past history of the vessel's flooding. The junior deck-hand did not appreciate the significance of the overboard discharge from the shelter deck and did not notify the master. Despite the list, no round of the shelter deck was made to ensure that everything was in order before both masters sailed on the JO MARC. The magnetic latch on the refrigerator door permitted the door to swing open when the vessel listed heavily to starboard. The position of the refrigerator door and the dry store door was such that the two door handles became interlocked when the refrigerator door swung open due to the heavy list. The interlocking of the door handles effectively trapped the chief engineer in the dry store. The owners carried out modifications to the MARWOOD which affected her stability but did not notify the regulatory authority. Despite several steamship inspections, the obvious modifications went unnoticed and Steamship Inspection Certificates were issued. The nets which were stowed on the after shelter deck were not secured and fell into the steering compartment, completely blocking the emergency exit access from the accommodation. 3.2 Causes While the vessel lay alongside the wharf unattended with a starboard list, the overhang of the spare trawl doors on the starboard side caught under the wale of the wharf on a rising tide, causing the MARWOOD to list heavily to starboard. The shelter deck then flooded by way of the open overboard drain valve to the conveyor well and of the non-watertight covers/openings to the holds/tanks. Downflooding of the spaces below deck ensued until the MARWOOD lost reserve buoyancy and sank. The fact that the overboard discharge was below the load waterline and that the conveyor well drain line was not fitted with a non-return valve, coupled with the practice of leaving the overboard drain valve open when not in use, contributed to the occurrence. 4.0 Safety Action 4.1 Action Taken 4.1.1 Vessels Sinking at Dock Following this accident, the Canadian Coast Guard (CCG) issued Ship Safety Bulletin No. 18/94, entitled Vessels Sinking at Dock, highlighting the need for vigilance and the precautions to be taken when leaving a vessel unattended. 4.1.2 Seaworthiness and Vessel Stability The TSB has apprised Transport Canada - Marine, via two Marine Safety Advisories, of the potential flooding hazards associated with the fish conveyor well discharge valve arrangement. Transport Canada - Marine was also advised to consider re-assessing the requirements respecting fish tanks to ensure structural integrity and seaworthiness of fishing vessels that carry fish in tanks. 4.2 Safety Concern 4.2.1 Ship Inspection Procedures and Practices Under Section 377-2 of the Canada Shipping Act, it is the responsibility of the owner or master of the vessel to inform the CCG of any alterations or modifications made to the vessel which may affect compliance with regulations. Often, however, fishermen and fishing vessel operators are not aware that the modifications they have made may affect the stability of the vessel and that the vessel, therefore, may not be in compliance. As a consequence, they do not notify the CCG of the modifications. Moreover, the Board has noted that, in some instances, potential safety hazards resulting from modifications to vessels have not been detected during subsequent periodical inspections by CCG marine surveyors. To address these shortcomings in the reporting and inspection of modifications to fishing vessels, the Board previously recommended that: The Department of Transport explore means to ensure that structural modifications and the addition of weight items are recorded and accounted for in re-assessing the stability of small fishing vessels. In its response to this recommendation, the Department of Transport reiterated the requirements of the Canada Shipping Act, Section 377-2, and of the Small Fishing Vessel Inspection Regulations. There was no indication that any new action would be taken to ensure that structural modifications were recorded and that vessel stability was reassessed. The vessel in this occurrence, the MARWOOD, had undergone substantial modifications in 1987 that affected her stability. The MARWOOD subsequently underwent six annual and one quadrennial safety inspections; apparently, the CCG marine surveyors did not note these modifications. The original Stability Booklet would have been inaccurate over the six-year period, yet Steamship Inspection Certificates continued to be issued. The Board recognizes that the circumstances surrounding the MARWOOD occurrence existed before the TSB issued Recommendation M94-32. However, initiatives have not been taken with respect to CCG inspection procedures and practices for the effective monitoring of vessel modifications. Therefore, the Board is concerned that existing unsafe conditions will continue to put the lives of fishermen and the safety of fishing vessels at risk. As such, the Board stands by the intent of Recommendation M94-32, that better means are needed to ensure that structural modifications and the addition of weight items are recorded and accounted for in reassessing the stability of small fishing vessels.